Missouri University of Science and Technology

Scholars' Mine
International Conference on Case Histories in
Geotechnical Engineering

(2013) - Seventh International Conference on Case

Histories in Geotechnical Engineering

Apr 29th - May 4th

High Capacity Rock Socketed Piles in Scotland,

UK
Peter Boyd
AECOM Ltd., United Kingdom

Baran Ozsoy
Zirve University, Turkey

Follow this and additional works at: http://scholarsmine.mst.edu/icchge

Part of the Geotechnical Engineering Commons
Recommended Citation
Peter Boyd and Baran Ozsoy, "High Capacity Rock Socketed Piles in Scotland, UK" (April 29, 2013). International Conference on Case
Histories in Geotechnical Engineering. Paper 19.
http://scholarsmine.mst.edu/icchge/7icchge/session02/19

This Article - Conference proceedings is brought to you for free and open access by Scholars' Mine. It has been accepted for inclusion in International
Conference on Case Histories in Geotechnical Engineering by an authorized administrator of Scholars' Mine. This work is protected by U. S. Copyright
Law. Unauthorized use including reproduction for redistribution requires the permission of the copyright holder. For more information, please contact
scholarsmine@mst.edu.

HIGH CAPACITY ROCK SOCKETED PILES IN SCOTLAND, UK

Peter Boyd
AECOM Ltd, 63 - 77 Victoria Street,
St Albans, AL1 3ER, Hertfordshire,
United Kingdom
peter.boyd@aecom.com

Dr. Baran Ozsoy

(formerly AECOM Ltd) Zirve University,
Kizilhisar Kampusu, 27260, Gaziantep,
Turkey
baran.ozsoy@zirve.edu.tr

ABSTRACT
The New Victoria Hospital in Glasgow, Scotland provides a 30,000m Ambulatory Care and Diagnostic (ACAD) facility. The
ground conditions beneath the site comprise a variable thickness of Made Ground overlying Glaciomarine Deposits and Glacial Till
which in turn rests on the Carboniferous bedrock. However the solid geology is complex and interrupted by a number of faults. The
Carboniferous rocks under the site include a number of coal seams, although it is thought that none of these seams has been worked at
this location. Due to the relatively low strength of the glacial soils, it was necessary to support the building on piles socketed into the
bed rock. However the various coal layers could not be relied upon to provide adequate end bearing capacity and therefore piles were
designed to be supported solely on side resistance from the rock socket. In order to investigate pile load bearing capacity and to
differentiate end bearing from shaft friction capacity, a number of preliminary and working load tests were carried out with one pile
particularly using a soft toe system. The results of load tests revealed an ultimate shaft friction capacity value in the rock socket of
approximately 1.3MPa (189psi).

THE DEVELOPMENT
The site forms part of a former battleground of the Battle of
Langside which was fought on 13 May 1568. Some 300 men
were killed during the battle although it is recorded that the
conflict lasted just 45 minutes. The site appears to have
remained as open fields for the many years thereafter and the
first development on the site occurred in the late 1800s when a
road was established through the site and a school, later
referred to as Queens Park Secondary School, was built on the
south western corner of the site. It was at this time that the
Victoria Infirmary (named after Queen Victoria) was built on
the land opposite. Some houses were later built on the site
north of the school, but by the late 1970s these had been
demolished and by 2005 the one remaining school building
had fallen into disrepair.
By this time the local health authority, the NHS Greater
Glasgow Health Board, had identified the need for a new
Ambulatory Care and Diagnostic (ACAD) Hospital to be built
on the site. The proposed new 100million development
comprised a four storey building and a semi-basement to be
built on the now largely derelict land. The development site
extended to approximately 0.3 hectares in area.

Paper No. 2.08

Under a PPP/PFI finance arrangement, the ultimate client

appointed the Canmore Consortium to deliver the project.
Balfour Beatty, a consortium member, was appointed to
design and construct the new building and in turn AECOM
was appointed as Balfour Beattys civil, structural and
geotechnical designer. The subsequent piling work was carried
out by Stent Foundations Ltd. The architectural design of the
project was performed by HLM Architects.
The preliminary structural design of the proposed buildings
called for the construction of nearly 400No 600mm and
750mm diameter bored cast in-situ piles socketed into the
rock. The principles of the pile design with initial load
capacities were produced by AECOMs Geotechnical Group.
The pile performance criteria were established in consultation
with AECOMs structural designers to ensure the permissible
settlement and angular distortion limits of the proposed
structure were not exceeded. The final pile design was
developed by the piling contractor Stent Foundations in line
with the recommendations in the geotechnical interpretative
report prepared by AECOM.
This paper describes the work undertaken to investigate the
site and design the pile foundations, focusing particularly on
the rock socket design and pile load testing.

GEOLOGICAL HISTORY
Various phases of ground investigation were carried out at the
site and these were latterly supplemented by a further
investigation focused on the proposed development.

The published geological map of the area shows the Dechmont

fault running through site close to the northern boundary. The
strata either side of the fault are quite distinct with the Middle
(Productive) Coal Measures subcropping north of the fault and
the Upper Limestone Formation subcropping to the south on
the downthrow side.

Superficial Geology
The published geological map of the area indicates the
majority of the site to be underlain by Quaternary drift
deposits of the late Devensian stage.
Table 1. Published Superficial Geology of the Area
Formation
Paisley Formation
Wilderness Till

Provenance

Age (years bp)

Marine
Glacial

11,500 -13,500
13,500-27,500

The Paisley Formation comprises glaciomarine deposits

typically manifest as sands, silts and clays. Its thickness in the
Clyde valley is typically around 5m but is locally absent. The
Wilderness Till is a glacial till comprising boulders and stones
in a hard to stiff sandy silty clay matrix. Its thickness is highly
variable and significant thicknesses may be found in infilled
glacial valleys.

However drilling and interpretation of the mining geology

during one of the early phases of ground investigation cast
doubt on the accuracy of the published map and at least two
further faults were inferred from the new data. The reinterpretation of the mapped geology indicated that the
Limestone Coal Formation (which includes numerous
workable coal seams) may underlie much of the northern half
of the site. South of the conjectured fault The Passage Group
may be present and beyond a further fault near the southern
end of the site the Upper Limestone is conjectured lie at
subcrop. Whilst both the Upper Limestone and Passage
Group include some potentially workable coal seams, they
are not thought to have been worked. Furthermore there are no
recorded workings in the Middle Coal Measures
immediately underlying the site.

Solid Geology
The published geological map shows the solid strata in the
area to form the following succession:
Table 2. The Published Solid Geology of the Area
Formation

Description

Significant Seams

Middle Coal
Measures

Sandstones,
siltstones and
mudstones with
numerous coal
seams
Mainly sandstones
with fireclays and
thin mudstones and
coals
Sandstones with
mudstones, thin
limestones and
coals
Sandstones,
siltstones and
mudstones with
numerous coals
and ironstones

Glasgow Main
Coal. Humph
Coal, Glasgow
Splint Coal, Virgin
Coal
No named seams

Passage Group

Upper Limestone
Formation

Limestone Coal
Formation

Paper No. 2.08

Fig.1 Site Plan and Borehole Locations

GROUND INVESTIGATION DATA

Lyoncross
Limestone,
Lyoncross Coal,
Index Limestone
Ashfield Coking
Coal, King Coal,
Possil Main Coal,
Pollock Stone
Coal. Glasgow
Shale Coal,
Jubilee Coal

Stratigraphy
The detailed ground investigation revealed a variable
thickness of superficial deposits overlying rockhead, the depth
to rockhead from surface varying from 6 m to 20.4m. The
majority of the site was found to be underlain by made ground
and this is turn was found to overlie a highly variable
succession of Glaciomarine deposits comprising loose sand
and silt overlying soft and occasionally laminated clays. The
Glaciomarine deposits were often found to rest on the Glacial
Till (stiff boulder clay) but the thickness of the Till was

generally limited and in some cases appeared to be absent. The

Glaciomarine deposits were deemed unsuitable for support of
the proposed building and therefore the design solution was to
support the new hospital on piles socketed into the bedrock.

As discussed in the next section, the RQD is also a significant

factor in calculating the pile rock socket capacity when direct
measurement of rock mass factor is not available. As can be
seen the majority of results are below 50% (Fig. 3).
Rock Quality Designation, RQD (%)
0

Design parameters

(1)

quc=24 Is(50)

Other researchers such as Rusnak and Mark (2000) have

derived strata-specific correlations for Carboniferous rocks
and they propose a similar correlation coefficient for these
rocks of around 21.
The data obtained for the ACAD site are plotted on Fig.2 and
the design line is drawn preferentially through the quc data.
Unconfined Compressive Strength, UCS (MPa)
40

40

60

80

100

The strength of the bedrock was determined from uniaxial

unconfined compression and point load index testing. The
point load index tests were carried out on cores in either an
axial or diametrical orientation. Generally it is the axial tests
that are correlated with unconfined compressive strength test
data. The point load index was converted to unconfined
compressive strength using the relationship proposed by Broch
and Franklin (1972) who established that a reasonable
correlation exists between the uniaxial compressive strength
(quc) and the point load strength index (Is(50)), where;

20

80

120

160

200

10

15

20

) 25
L
G
B
(m30
h
t
p
e
D
35

40

45

50
240

55
1

60
2

Fig.3 RQD Measurements on Rock Core

APPROACH TO PILE DESIGN

)
m
(
d
a 4
e
h
k
c
o
R
w 5
lo
e
b
h
t
p 6
e
D

The rock strata underlying the site comprised a succession of

mudstones, siltstones, sandstones as well as a number of coal
seams. Whilst it is known that the coals seams have not been
worked (generally they are too thin and too deep to be of
economic value) the coal seams themselves are significantly
weaker than the surrounding rock.

8
Unconf ined Com pressive Test
Point Load Test - Axial

Point load tests on coal indicated unconfined compressive

strengths as low as 0.5MPa (73psi), whereas the surrounding
rock has a characteristic unconfined compressive strength of at
least 20MPa (2900psi).

9
Point Load Test - Coal
Design Line
10

Fig.2 Unconfined Compressive Strengths Derived from UCS

tests and Point Load Tests

Paper No. 2.08

Given that the occurrence and depth of coal layers beneath the
site has been shown by the ground investigation to be
unpredictable (largely as a the result of significant local
faulting), there was considered to be a significant risk of a pile
tip bearing on, or just above, a weak coal seam. End bearing
resulting from bearing onto coal is only a small proportion of

that resulting from bearing on competent rock. Therefore it

was felt prudent, for preliminary design purposes, to ignore
the contribution to pile capacity from end bearing.
It was important therefore to establish a reliable and not
unduly conservative value for the load capacity which could
be derived from the rock socket side friction alone. Various
methods of calculating the capacity of the rock socket were
considered, the majority of which relate rock socket capacity
to unconfined compressive strength. However other studies
have established that further factors such as joint spacing and
roughness of the socket may have a significant influence on
the mobilised rock socket resistance - see Haberfield and
Collingwood (2006).

UK normal practice is to resort to well-established guidance

such as Pile Design and Construction Practice by
Tomlinson and Woodward (2008) and previous editions of the
same text. In turn Tomlinson and Woodward (2008) and
previously Tomlinson (1994) cite a number of methods of
calculating pile rock socket capacity, namely Horvarth (1978),
Rosenberg and Journeaux ( 1976) and Williams and Pells
(1981).
All the above, as reported by Tomlinson and Wood (2008),
relate the ultimate rock socket bond strength (fs) to the
unconfined compressive strength using the following
equation:
fs = ..quc

Given the wide range of values which could be derived from

theoretical calculations plus the inherent uncertainty in
determining an accurate value for the unconfined compressive
strength of the rock, it was felt important to undertake pile
load testing to establish the operational in situ strength of the
rock socket.

PILE PERFORMANCE CRITERIA

(2)

where:
=
=
quc =

Reduction factor relating to quc

Correction factor related to discontinuity spacing in
the rock mass.
Average unconfined compressive strength of the
rock over the length of the rock socket

The piling work and load testing performance criteria were set
out in accordance with the Institution of Civil Engineers
Specification for Piling and Embedded Retaining Walls
(SPERW) dated 1996.

Whilst the Williams and Pells (1981) method gives the highest
value of the other two methods cited by Tomlinson assume
an value of unity. Therefore the approach by Williams and
Pells (1981) is likely to be more conservative than the two
other methods when considering highly fractured rocks.

The acceptance criteria for the test piles selected for the
project were as follows:

The mass factor j is defined as the ratio of the elastic modulus

of the rock mass to that of the intact rock.

a)

Maximum first cycle settlement at safe working load

(SWL) not more than 10mm.
b) Maximum settlement at 150% of SWL not more than
20mm.
c) Residual settlement after second cycle of loading to
150% of working load, to be not more than 50% of
permitted settlements at SWL.

Ideally this is measured using geophysical techniques or

loading tests, but can be estimated from a knowledge of the
discontinuity spacing and/or RQD from the recovered rock
cores.

It was also stipulated that the pile foundations should be

designed so that the differential movement between adjacent
pile caps would not exceed 1 in 500.
A factor of safety of at least 2.0 on skin friction for
compression loads and a factor of safety of 3.0 on tension
loads was required the latter principally due to temporary
loading combinations during the construction stages of the
structural work.

CALCULATION OF ROCK SOCKET CAPACITY FROM

GROUND INVESTIGATION DATA
In the UK the British Standard BS 8004 Foundations (1986)
(now withdrawn and superseded by Eurocodes) provides little
guidance on the design of rock sockets for piles and common

Paper No. 2.08

Fig.4 Reduction factors for discontinuities in rock mass (after

Williams and Pells, 1981)

A relationship between RQD and mass factor j was proposed

by Hobbs (1975) as follows:
Table 3. Relationship between RQD and Rock Mass factor j
RQD (%)

Fracture Frequency/m

Mass factor j

0-25
25-50
50-75
75-90
90-100

15
15-8
8-5
5-1
1

0.2
0.2
0.2-0.5
0.5-0.8
0.8-1.0

On the basis of the foregoing the following parameters were

derived:
= 0.10
quc = 20MPa (2900psi)
= 0.65 (RQD assumed to be 50% or less)
Hence fs = 1.3MPa (189psi)
A useful historical review of the various factors derived by
various researchers, as well as their own recommendations,
has been presented by Kulhawy et al. (2005). Various others
methods of calculating the capacity of rock sockets are
available. By and large these take the form as follows:
fs/Pa = c.(quc/Pa)n

(3)

Where Pa = atmospheric pressure and C and n are empirical

factors.
A particular relationship of this type proposed by Fleming et
al. (1992) is as follows
fs/Pa = 1.3(quc/Pa)0.5

(4)

Table 4. Published Relationships between Unconfined

Compressive Strength and Rock Socket Bond Strength
Reference

fs
(MPa)
1.71

Rosenberg and
1.09
0.52
Journeaux
(1976)
Horvath (1978)
1.04
0.50
1.47
Horvath &
0.650.50
0.92
Kenney (1979)
0.78
Williams et al.
1.84
0.37
1.31
(1980)
Rowe &
1.42
0.50
2.01
Armitage (1984)
Carter &
0.63
0.50
0.89
Kulhawy (1988)
Fleming et al.
1.30
0.50
1.84
(1992)*
Zhang &
1.26
0.50
1.78
Einstein (1999)
Prakoso (2002)
0.98
0.50
1.39
Kulhawy et al
1.00
0.50
1.41
(2005)
(*)
Not included in Kulhawy et al. data

(psi)
248

213
133
190
292
129
267
258
202
205

For the ACAD site (assuming quc= 20MPa), the above

methods yield shaft friction values in the range of 0.89MPa
(129psi) to 2.01MPa (292psi)., the highest value being given
by Rowe and Armitage (1984).
The above data contrasts with local practice in the Glasgow
area was to assume much lower values for shaft friction
presumed values being typically 0.25MPa (36psi) for
mudstone, 0.5MPa (73psi) for siltstone and 0.75MPa (109psi)
for sandstone and limestone with values being associated
with rock type rather than directly correlated with rock
strength.

Additionally Fleming at a.(1992) stipulate that the above

relationship is only acceptable for use in rock sockets where
the shaft is sufficiently rough to ensure full keying of the
concrete and the host rock. Furthermore they caution that fs
should not exceed 5% of the concrete strength.

In view of this disparity, it was therefore felt to be imperative

to verify any higher values by means of pile load testing.

Since fs for the pile designs derived using the approach of

Williams and Pells (1981) does not exceed either of these
values, it was considered that the unit skin friction value
adopted was appropriate

1st Preliminary Test Pile with the Soft Toe Feature (TP1)

A comparison of the values derived from various published

methods is provided in Table 4.

PILE LOAD COMPRESSION TESTS

The pile design called for a 750mm (2 6) diameter pile with

a 3m (9) long rock socket. The nominal safe working load
(SWL) of the test pile was calculated to be 4200kN (944kipf).
The achieved peak test load was 9450kN (2124kipf).
The test pile location was deliberately selected to be close to a
borehole location in order to facilitate a correlation the soil
parameters to the observed pile behaviour. The ground
conditions at the test pile location were recorded as 1.2m of

Paper No. 2.08

Made Ground overlaying 7.5m of Glaciomarine Deposits and

Glacial Till. A 0.5m (20) thick coal layer was recorded by the
piling contractor and the top of the bedrock was observed to
be 9.2m below the ground surface.
A 300mm (12) thick soft toe made of polystyrene was placed
beneath the reinforcement cage..

such discontinuities may have also contributed to an overall

reduction in the vertical rock stiffness measured.
The difficulty in accurately predicting the behaviour of rock
sockets under load was also highlighted by Pells (1999) who
found that the Youngs modulus was often highly variable
even when tests were carried out in the same rock mass,

The load test was a maintained load test carried out in

accordance with ICE Specification for Piling which includes
three loading-unloading loops and a number of load holding
stages.
The load settlement relationship for the pile test undertaken
with a soft toe is shown below.
Load (kN)
0

2000

4000

6000

8000

10000

0
10

TP1

20

)
m
m
(t 30
n
e 40
m
e
v
o 50
M
60
70
80

Fig.5 Pile Load Compression Test Results for TP1 (Soft Toe)
The test was taken to a maximum load of 9450kN (2125kip-f)
which equates to 225% SWL. It was not possible to maintain
the load beyond this point and in view of the soft toe it was
felt advisable for health and safety reasons to terminate the
test when settlement reached 75mm (3). This settlement
corresponds to 10% of the pile diameter and this is in itself an
arbitrary definition of pile failure.
A back analysis of the load-settlement data was undertaken
using the CEMSOLVE program using the method derived by
Fleming (1992). This analysis suggested that higher ultimate
load might have been achievable albeit at very large
settlements. However this could not be confirmed with the
method of testing adopted.
The predicted elastic shortening obtained from CEMSOLVE
using E=30GN/m2 gave values close to the settlement recovery
at the end of the test of around 10mm. However the actualload settlement response was much softer than that might
have been predicted at the outset using CEMSET. Specifically
the shaft flexibility factor Ms was back calculated by curve
fitting and this yielded a value of around 0.005 a value
associated with soft soils. Ms is in fact the tangent slope at the
original of the hyberbolic function representing shaft friction.
The reasons for this disparity are not clear but it is possible
that rock discontinuities and overall roughness of the socket
will have increased the load bearing capacity of the socket but

Paper No. 2.08

Fig.6 The Soft Toe Detail for the First Test Pile (TP1)
2nd Preliminary Test Pile without the Soft Toe Feature (TP2)
The second test pile was constructed without the soft toe
feature in order to make a comparison with TP1. The
designated working load for TP2 was 4200kN (944kipf) and
the target peak test load was 10500kN (2360kipf). The pile
settled 8.2mm (0.32) at the peak test load and a residual
settlement of 3.2mm (0.12) was observed at the end of the
test.
The most significant difference between TP1 and TP2 was the
stiffer pile settlement response observed during the second
preliminary test. It is interpreted that due to the relatively short
rock socket length in both cases, the contribution of pile endbearing capacity to the overall pile capacity was more evident
in the second test and some significant end-bearing capacity
might have mobilized before the pile test mobilized the full
shaft capacity.
Load (kN)
0

2000

4000

6000

8000

10000

12000

0
1

TP2

)
m
(m
t
n
e
m
e
v
o
M

3
4
5
6
7
8
9
10

Fig.7 Pile Load Compression Test Results for TP2

In summary both tests were concluded to be satisfactory and

the piling contractor commenced the site works with no
change in the pile design philosophy.

The Working Load Tests for QA/QC Purposes (TP3 and TP4)
Two other test piles were selected for QA/QC purposes and
maintained load tests were performed on these contract piles
up to 150% of their SWL.. The test pile diameters for TP3 and
TP4 were 750mm and 600mm; respectively. The performance
of these two contract piles were also considered satisfactory as
they satisfied the structural performance criteria set out at the
piling specification.
Load (kN)
0

1000

2000

3000

4000

5000

BOND STRENGTH COMPARISON WITH PUBLISHED

DATA
The measured average bond strength from the compressive
pile load with the soft toe test was 1.3MN/m2 (1.89psi). This
matches closely with the predicted value derived using the
William and Pells (1981) approach which takes account of
both unconfined compressive strength and the rock mass
factor.
Taking an assumed unconfined compressive strength of
20MPa (2900psi), the measured bond stress was equivalent
0.065qu. This lies close to, but slightly below the trend line
produced by Long (2000) for Carboniferous rocks in Ireland .
When compared to the various predictive methods cited by
Kulhawy et al (2005) the observed strength is given by the
following relationship:

TP3
TP4

fs/Pa = 0.92 (quc/Pa)0.5

(5)

)
m
m
(t 2
n
e
m
e
v
o 3
M
4

Fig.8. Pile Load Compression Test Results for TP3 and TP4

Summary of Pile Compression Load Test Results

In total, four compression load tests were performed in this
project. The test pile diameters were 750mm for TP1, TP2 and
TP3. The remaining load test was performed on a 600mm (2)
diameter pile..

Fig.9 A View of the Piling Works Performed by Stent

PILE LATERAL LOAD TESTING

Table 5. Summary of Pile Loads and Settlements during the

Compression Load Testing Programme
Pile No

WL (kN)

PTL (kN)

Settlement
@ WL
(mm)

TP1
4200
9450
7.17
TP2
4200
10500
2.36
TP3
3000
4500
1.77
TP4
2500
3750
1.76
(*)
WL: Working Load / PTL: Peak Test Load

Settlement
@ PTL
(mm)
75.39
8.20
3.30
2.90

Note: The pile settlements were measured at the pile heads by

averaging the values recorded electronically on four dialgauges.

Paper No. 2.08

In the long term the lateral loads to be resisted by the piles

were estimated to be very small as the pile caps will be
restrained by means of stiff ground slabs spanning in both
directions. During the design process of the superstructure it
was revealed that the overall construction time could be
significantly reduced if the erection of the structural frames
could take place, without waiting the casting of the ground
slabs. This would require the individual piles to be designed to
resist significantly higher lateral loads, as they would be
subject to lateral wind loads acting on the structural frame
during the temporary construction stages.
In order to measure the lateral behavior of the contract piles
two lateral load tests were performed on 600mm (LTP1) and
750mm (LTP2) piles. The working lateral load was
determined to be 275kN whereas the peak test load was
330kN in both cases.

Load (kN)
0

50

100

150

200

250

300

350

400

LTP1
LTP2
1

)
m
(m
t 2
n
e
m
e
v
o 3
M

The preliminary design of the rock sockets was based on a

conservative assessment of the rock strength and so the soft
toe pile testing allowed a higher rock socket bond strength
and hence a more economic pile design to be adopted. The
lateral load test also helped the structural engineer and the
client to shorten the overall construction period.

ACKNOWLEDGEMENTS
The authors are grateful to AECOM, Balfour Beatty and Stent
Foundations for providing the necessary data to prepare this
paper.

REFERENCES
Fig.10 Pile Lateral Load Test Results

Broch, E. and J.A Franklin (1972).The Point Load Strength

Test. Int. Journal Rock Mech. Min. Sci 9, pp. 669-697.

A summary of the lateral load test results is as follows:

Table 6. Summary of Pile Behaviour during the Lateral Load
Testing Programme

BSI (1986). BS8004.Code of Practice for Foundations. British

Standards Institute. London.

Pile No

Deflection
@ PTL
(mm)

Carter, J.P. and.F.H. Kulhawy (1988). Analysis and design of

drilled shaft foundations socketed into rock. Report EL-5918.
Palo Alto: Electric Power Research Institute.

4.59
1.79

Fleming, W.G.K. (1992). A new method for single pile

settlement and analysis. Geotechnique, 42, No 3 pp 413-425.

WL (kN)

PTL (kN)

Deflection
@ WL
(mm)

LTP1
275
330
2.88
LTP2
275
330
1.25
(*)
WL: Working Load / PTL: Peak Test Load

The resultant deformations were considered satisfactory by the

structural engineers and the piling works were commenced
accordingly. The lateral load tests resulted in significant
savings in terms of overall construction time and in turn
project budget.

Fleming, W.G.K., A.J. Weltman, M.F. Randolph and

W.K.Elson (1992). Piling Engineering, John Wiley & Sons,
Inc., New York, N.Y.
Haberfield, C. and B. Collingwood (2006). Rock-socketed
pile design and construction: a better way? Proc. ICE,
Geotechnical Engineering, Vol. 159, No.3, July 2006 , pp.
207217.

CONCLUSIONS
The predicted bond strength in a rock socket is usually
estimated from the unconfined compressive strength. However
the accuracy of such an approach is dependent upon obtaining
a representative value of the rock strength. At the Victoria
ACAD site the relationship proposed by Williams et al. (1980)
and also Williams and Pell (1981) gave the closest
approximation to the value which was later verified by the soft
toe pile testing. The method proposed by Kulhawy et al.
(2005) also gave a good approximation.
Although a number of load tests were performed for this
project, it will be necessary to carry out further research and
more rock socket load tests in Carboniferous rocks in the
Glasgow area before more widely applicable design guidance
can be formulated. It is certainly vital to carry out preliminary
load tests particularly as major uncertainties exist in the
prediction of rock socket behaviour.

Paper No. 2.08

Hobbs, N.B. (1975). Review Paper Rocks. Proc. of the

Conf. on Settlement of Structures. British Geotechnical
Society. Pentech Press, pp. 579-610.
Horvath, R.G. (1978). Field load test data on concrete-torock bond strength for drilled pier foundations. Publication
78-07, Toronto: Univ. of Toronto.
Horvath, R.G. and T.C. Kenney (1979) Shaft resistance of
rock-socketed drilled piers. In Symposium on Deep
Foundations, Atlanta, Oct. 1979, ed. F.M. Fuller, pp. 182-214.
New York: ASCE.
Horvath, R.G., T.C. Kenney and P. Kozicki (1983). Methods
of improving the performance of drilled piers in weak rock.
Canadian Geotech. J. 20, No 4, pp. 758-772.
Kulhawy, F.H., W.A. Prakoso and S.O. Akbas (2005)
Evaluation of Capacity of Rock Foundation Sockets. Alaska

Rocks 2005, The 40th U.S. Symp. on Rock Mechanics

(USRMS): Rock Mechanics for Energy, Mineral and
Infrastructure Development in the Northern Regions, held in
Anchorage, Alaska, June 25-29, 2005.
Long, M. (2000).Skin friction for piles socketed in hard
rock. Proc. GeoEng. 2000, Melbourne, Australia, November
2000.
Pells, P.J.N. (1999). State of Practice for the Design of
Socketed Piles in Rock. Proc. 8th Australia New Zealand
Conf. on Geomechanics: Consolidating Knowledge. Barton,
ACT: Australian Geomechanics Society, pp. 307-327
Prakoso, W.A. (2002). Reliability-based design of
foundations in rock masses. PhD Dissertation. Ithaca: Cornell
University.
Rosenberg, P. and N.L. Journeaux (1976). Friction and end
bearing tests on bedrock for high capacity socket design.
Canadian Geotech. J. 13(3), pp. 324-333.
Rowe, R.K. and H.H. Armitage (1984). Design of piles
socketed into weak rock. Report GEOT-11-84. London:
Univ. of Western Ontario.
Rowe, R.K.and H.H. Armitage (1987). A design method for

Paper No. 2.08

drilled piers in soft rock. Canadian Geotech. J.. 24(1), pp.

126-142.
Rusnak, J.and C. Mark (2000). Using the point load test to
determine the uniaxial compressive strength of coal measure
rock. In: Peng, S.S., Mark, C., eds. Proc.19th International
Conf. on Ground Control in Mining. Morgantown, WV: West
Virginia University, pp. 362-371.
Tomlinson, M.J (1994). Pile Design and Construction
Practice. 4th Edition. Taylor &Francis. London.
Tomlinson, M.J and J. Woodward (2008). Pile Design and
Construction Practice. 5th Edition. Taylor &Francis. London.
Williams, A.F., I.W. Johnston and I.B. Donald (1980).
Design of socketed piles in weak rock. In Structural
Foundations on Rock, ed. P.J.N. Pells, pp. .327-347.
Rotterdam: Balkema.
Williams, A. F. and P.J.N Pells (1981). Side Resistance Rock
Sockets in Sandstone, Mudstone and Shale, Canadian
Geotechnical J., 18, pp 502-513.
Zhang, L. and H.H. Einstein (1998). End bearing capacity of
drilled shafts in rock. J. Geotech. Eng.(ASCE), 24(7),
pp.574-584.